Can artificial intelligence (AI) be used in judicial processes to reduce the pendency of cases? In response to this unstarred question in the Lok Sabha during the first part of the Budget session of Parliament, Law Minister Kiren Rijiju said that while implementing phase two of the eCourts projects, under operation since 2015, a need was felt to adopt new, cutting edge technologies of Machine Learning (ML) and Artificial Intelligence (AI) to increase the efficiency of the justice delivery system.

“To explore the use of AI in judicial domain, the Supreme Court of India has constituted Artificial Intelligence Committee which has mainly identified application of AI technology in Translation of judicial documents; Legal research assistance and Process automation,” Mr. Rijiju stated.

Several law firms are now keen try out new technologies for a quick reference on judicial precedents and pronouncements on cases with similar legal issues at stake. Mumbai-based Riverus, a “legal tech” firm, has developed ML applications that peruse troves of cases, “understand” them, and parse cases that are similar in content — very much like a human expert would do — in a fraction of the time.

One of the applications, said Dipankar Bandopadhyay, a lawyer and the founder-director of the firm, is specific to tax law, and the other to preparing contracts. After “training” the application on a huge historical set of precedents, the application is capable of highlighting key points that are relevant in specific contracts.

“Say, you wanted to find the rulings of the court, or a specific judge, on black money — you can use the software to analyse thousands of previous cases and create a ‘judge analytics’. It’s quicker than having someone actually sit down and prepare a huge Excel sheet, which is too much work for a person,” Mr. Bandopadhyay said. “In contracts, I can parse it through the system and give you an X-ray sort of report on what points are missing, what are present.”

Over the course of the COVID-19 pandemic, the use of technology for e-filing, and virtual hearings has seen a dramatic rise. From the beginning of the lockdown in 2020 until January 8 this year, the Supreme Court of India emerged as a global leader by conducting 1,81,909 virtual hearings.

The High Courts (57.39 lakh cases) and the subordinate courts (1,08,36,087 cases) together have conducted 1.65 crore virtual hearings till November 30, 2021 according to Law Ministry data.

But the use of ML in India’s legal sphere has so far been restricted to automating back-end work, and is still a very long way from being used as a decision-making tool for the judiciary. Many of the judgments, particularly in the lower courts, are yet to be fully digitised, but experts like Mr. Bandopadhyay argue that going by global trends, greater adoption of these tools in the Indian legal system is inevitable.

SUVAS is a language learning application being used to translate judgments, and SUPACE, which can draft a legal brief, comprise the initiatives being undertaken in the Indian judiciary as part of incorporating ML-based applications, said Ameen Jauhar, who leads the Centre for Applied Law and Technology Research at the Vidhi Centre for Legal Policy.

Automated systems, controversially, were being used to decide bail applications in some parts of the United States, and other countries such as Estonia have incorporated AI and ML in a major way. But the Indian judicial system is generally “more conservative”, and a lot more work remained in making India’s legal data amenable to ML formats, he added.

Sanjay K. Chadha, managing partner of a Delhi-based law firm, BSK Legal, said AI and ML can be tried in tribunals where “there is no need for oral evidence and cross examination”.

“Consumer courts are an area where AI can be helpful. But in criminal cases where oral evidence and cross examination are key processes, we have to rely on regular human intervention,” Mr. Chadha added.